Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Large-scale neural recordings with single neuron resolution using Neuropixels probes in human cortex

Abstract

Recent advances in multi-electrode array technology have made it possible to monitor large neuronal ensembles at cellular resolution in animal models. In humans, however, current approaches restrict recordings to a few neurons per penetrating electrode or combine the signals of thousands of neurons in local field potential (LFP) recordings. Here we describe a new probe variant and set of techniques that enable simultaneous recording from over 200 well-isolated cortical single units in human participants during intraoperative neurosurgical procedures using silicon Neuropixels probes. We characterized a diversity of extracellular waveforms with eight separable single-unit classes, with differing firing rates, locations along the length of the electrode array, waveform spatial spread and modulation by LFP events such as inter-ictal discharges and burst suppression. Although some challenges remain in creating a turnkey recording system, high-density silicon arrays provide a path for studying human-specific cognitive processes and their dysfunction at unprecedented spatiotemporal resolution.

This is a preview of subscription content

Access options

Buy article

Get time limited or full article access on ReadCube.

$32.00

All prices are NET prices.

Fig. 1: Human neocortical neurons recorded using Neuropixels 1.0-S probes.
Fig. 2: A variety of waveform types and shapes recorded in human neocortex with Neuropixels probes.
Fig. 3: Classifying waveforms based on spatial and temporal features.
Fig. 4: Relationship of units to the LFP events and epileptiform discharges.

Data availability

The data is available for download at Dryad (https://doi.org/10.5061/dryad.d2547d840) upon publication.

Code availability

Open-source acquisition software, SpikeGLX Release v20201103-phase30 (http://billkarsh.github.io/SpikeGLX/), was used to record the neural data. Single-unit sorting was performed using Kilosort 3.0 (ref. 28) (https://github.com/MouseLand/Kilosort) as well as Phy2 (https://github.com/cortex-lab/phy) custom MATLAB code (version R2020a), and Python code in combination with open-source code from the Fieldtrip toolbox (http://www.fieldtriptoolbox.org/) was used for most of the analyses, with some code involving manual alignment available on GitHub (https://github.com/Center-For-Neurotechnology/CorticalNeuropixelProcessingPipeline). The burst suppression ratio was computed using an automated method65,81 (https://github.com/drasros/bs_detector_icueeg). Reconstruction of electrode locations and the manual tracing was done using the open-source, free software Blender (https://www.blender.org/), Mango (http://ric.uthscsa.edu/mango/) and MMVT (https://github.com/pelednoam/mmvt). Violin plots showing the distribution of the data based on kernel density estimate using ksdensity in several figures were produced using code by the Holger Hoffmann (2021) violin plot (https://www.mathworks.com/matlabcentral/fileexchange/45134-violin-plot) on the MATLAB Central File Exchange (retrieved 15 April 2021).

References

  1. Jun, J. J. et al. Fully integrated silicon probes for high-density recording of neural activity. Nature 551, 232–236 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  2. Jia, X. et al. High-density extracellular probes reveal dendritic backpropagation and facilitate neuron classification. J. Neurophysiol. 121, 1831–1847 (2019).

    PubMed  Google Scholar 

  3. Trautmann, E. M. et al. Accurate estimation of neural population dynamics without spike sorting. Neuron 103, 292–308.e4 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  4. Steinmetz, N. A. et al. Neuropixels 2.0: a miniaturized high-density probe for stable, long-term brain recordings. Science 372, eabf4588 (2021).

    CAS  PubMed  PubMed Central  Google Scholar 

  5. Ward, A. A. & Thomas, L. B. The electrical activity of single units in the cerebral cortex of man. Electroencephalogr. Clin. Neurophysiol. 7, 135–136 (1955).

    CAS  PubMed  Google Scholar 

  6. Verzeano, M., Crandall, P. H. & Dymond, A. Neuronal activity of the amygdala in patients with psychomotor epilepsy. Neuropsychologia 9, 331–344 (1971).

    CAS  PubMed  Google Scholar 

  7. Ulbert, I., Halgren, E., Heit, G. & Karmos, G. Multiple microelectrode-recording system for human intracortical applications. J. Neurosci. Methods 106, 69–79 (2001).

    CAS  PubMed  Google Scholar 

  8. Worrell, G. A. et al. High-frequency oscillations in human temporal lobe: simultaneous microwire and clinical macroelectrode recordings. Brain 131, 928–937 (2008).

    PubMed  Google Scholar 

  9. Nordhausen, C. T., Rousche, P. J. & Normann, R. A. Optimizing recording capabilities of the Utah Intracortical Electrode Array. Brain Res. 637, 27–36 (1994).

    CAS  PubMed  Google Scholar 

  10. Khodagholy, D. et al. NeuroGrid: recording action potentials from the surface of the brain. Nat. Neurosci. 18, 310–315 (2015).

    CAS  PubMed  Google Scholar 

  11. Paulk, A. C. et al. Microscale physiological events on the human cortical surface. Cereb. Cortex 8, 3678–3700 (2021).

    Google Scholar 

  12. Kleen, J. K. et al. Bidirectional propagation of low frequency oscillations over the human hippocampal surface. Nat. Commun. 12, 2764 (2021).

    CAS  PubMed  PubMed Central  Google Scholar 

  13. Truccolo, W. et al. Single-neuron dynamics in human focal epilepsy. Nat. Neurosci. 14, 635–641 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  14. Mukamel, R. & Fried, I. Human intracranial recordings and cognitive neuroscience. Annu. Rev. Psychol. 63, 511–537 (2012).

    PubMed  Google Scholar 

  15. Cash, S. S. & Hochberg, L. R. The emergence of single neurons in clinical neurology. Neuron 86, 79–91 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  16. Chari, A., Thornton, R. C., Tisdall, M. M. & Scott, R. C. Microelectrode recordings in human epilepsy: a case for clinical translation. Brain Commun. 2, fcaa082 (2020).

    PubMed  PubMed Central  Google Scholar 

  17. Amirnovin, R., Williams, Z. M., Cosgrove, G. R. & Eskandar, E. N. Experience with microelectrode guided subthalamic nucleus deep brain stimulation. Neurosurgery 58, ONS96-102 (2006).

    Google Scholar 

  18. Jamali, M. et al. Dorsolateral prefrontal neurons mediate subjective decisions and their variation in humans. Nat. Neurosci. 22, 1010–1020 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  19. Hochberg, L. R. et al. Neuronal ensemble control of prosthetic devices by a human with tetraplegia. Nature 442, 164–171 (2006).

    CAS  PubMed  Google Scholar 

  20. Hochberg, L. R. et al. Reach and grasp by people with tetraplegia using a neurally controlled robotic arm. Nature 485, 372–375 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  21. Stavisky, S. D. et al. Neural ensemble dynamics in dorsal motor cortex during speech in people with paralysis. eLife 8, e46015 (2019).

  22. Stavisky, S. D. et al. Speech-related dorsal motor cortex activity does not interfere with iBCI cursor control. J. Neural Eng. 17, 016049 (2020).

    PubMed  PubMed Central  Google Scholar 

  23. Willett, F. R., Avansino, D. T., Hochberg, L. R., Henderson, J. M. & Shenoy, K. V. High-performance brain-to-text communication via handwriting. Nature 593, 249–254 (2021).

    CAS  PubMed  PubMed Central  Google Scholar 

  24. Flesher, S. N. et al. Intracortical microstimulation of human somatosensory cortex. Sci. Transl. Med. 8, 361ra141 (2016).

  25. Venkataramani, V. et al. Glutamatergic synaptic input to glioma cells drives brain tumour progression. Nature 573, 532–538 (2019).

    CAS  PubMed  Google Scholar 

  26. Venkatesh, S. et al. Electrical and synaptic integration of glioma into neural circuits. Nature 573, 539–545 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  27. Varol, E. et al. Decentralized motion inference and registration of neuropixel data. In: 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 1085–1089 https://doi.org/10.1109/ICASSP39728.2021.9414145 (2021).

  28. Pachitariu, M., Steinmetz, N., Kadir, S., Carandini, M. & Harris, K. Fast and accurate spike sorting of high-channel count probes with KiloSort. Adv. Neural Inf. Process. Syst. 2016, 1–9 (2016).

    Google Scholar 

  29. Buzsáki, G., Anastassiou, C. A. & Koch, C. The origin of extracellular fields and currents—EEG, ECoG, LFP and spikes. Nat. Rev. Neurosci. 13, 407–420 (2012).

    PubMed  PubMed Central  Google Scholar 

  30. Gold, C., Henze, D. A. & Koch, C. On the origin of the extracellular action potential waveform: a modeling study. J. Neurophysiol. 95, 3113–3128 (2006).

    CAS  PubMed  Google Scholar 

  31. Lee, E. K. et al. Non-linear dimensionality reduction on extracellular waveforms reveals physiological, functional, and laminar diversity in premotor cortex. eLife 10, e67490 (2021).

    PubMed  PubMed Central  Google Scholar 

  32. Sun, S. H. et al. Analysis of extracellular spike waveforms and associated receptive fields of neurons in cat primary visual cortex. J. Physiol. 8, 2211–2238 (2021).

    Google Scholar 

  33. Zhu, S., Xia, R., Chen, X. & Moore, T. Heterogeneity of neuronal populations within columns of primate V1 revealed by high-density recordings. Preprint at https://www.biorxiv.org/content/10.1101/2020.12.22.424048v1.abstract (2020).

  34. Mitchell, J. F., Sundberg, K. A. & Reynolds, J. H. Differential attention-dependent response modulation across cell classes in macaque visual area V4. Neuron 55, 131–141 (2007).

    CAS  PubMed  Google Scholar 

  35. McCormick, D. A., Connors, B. W., Lighthall, J. W. & Prince, D. A. Comparative electrophysiology of pyramidal and sparsely spiny stellate neurons of the neocortex. J. Neurophysiol. 54, 782–806 (1985).

    CAS  PubMed  Google Scholar 

  36. Bartho, P. et al. Characterization of neocortical principal cells and interneurons by network interactions and extracellular features. J. Neurophysiol. 92, 600–608 (2004).

    PubMed  Google Scholar 

  37. Kaufman, M. T. et al. Roles of monkey premotor neuron classes in movement preparation and execution. J. Neurophysiol. 104, 799–810 (2010).

    PubMed  PubMed Central  Google Scholar 

  38. Swadlow, H. A. Fast-spike interneurons and feedforward inhibition in awake sensory neocortex. Cereb. Cortex 13, 25–32 (2003).

    PubMed  Google Scholar 

  39. Swadlow, H. A. Efferent neurons and suspected interneurons in motor cortex of the awake rabbit: axonal properties, sensory receptive fields, and subthreshold synaptic inputs. J. Neurophysiol. 71, 437–453 (2017).

    Google Scholar 

  40. Okun, M. et al. Diverse coupling of neurons to populations in sensory cortex. Nature 521, 511–515 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  41. Sakata, S. & Harris, K. D. Laminar structure of spontaneous and sensory-evoked population activity in auditory cortex. Neuron 64, 404–418 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  42. Dearnley, B., Dervinis, M., Shaw, M. & Okun, M. Stretching and squeezing of neuronal log firing rate distribution by psychedelic and intrinsic brain state transitions. Preprint at https://www.researchgate.net/publication/354091542_Stretching_and_squeezing_of_neuronal_log_firing_rate_distribution_by_psychedelic_and_intrinsic_brain_state_transitions (2021).

  43. Ching, S., Purdon, P. L., Vijayan, S., Kopell, N. J. & Brown, E. N. A neurophysiological–metabolic model for burst suppression. Proc. Natl Acad. Sci. USA 109, 3095–3100 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  44. Sheth, S. A. et al. Human dorsal anterior cingulate cortex neurons mediate ongoing behavioural adaptation. Nature 488, 218–221 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  45. Janca, R. et al. Detection of interictal epileptiform discharges using signal envelope distribution modelling: application to epileptic and non-epileptic intracranial recordings. Brain Topogr. 28, 172–183 (2015).

    PubMed  Google Scholar 

  46. Keller, C. J. et al. Heterogeneous neuronal firing patterns during interictal epileptiform discharges in the human cortex. Brain 133, 1668–1681 (2010).

    PubMed  PubMed Central  Google Scholar 

  47. Lewis, L. D. et al. Local cortical dynamics of burst suppression in the anaesthetized brain. Brain 136, 2727–2737 (2013).

    PubMed  PubMed Central  Google Scholar 

  48. Kaufman, M. T., Churchland, M. M. & Shenoy, K. V. The roles of monkey M1 neuron classes in movement preparation and execution. J. Neurophysiol. 110, 817–825 (2013).

    PubMed  PubMed Central  Google Scholar 

  49. Peyrache, A. et al. Spatiotemporal dynamics of neocortical excitation and inhibition during human sleep. Proc. Natl Acad. Sci. USA 109, 1731–1736 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  50. Hodge, R. D. et al. Transcriptomic evidence that von Economo neurons are regionally specialized extratelencephalic-projecting excitatory neurons. Nat. Commun. 11, 1172 (2020).

    CAS  PubMed  PubMed Central  Google Scholar 

  51. Boldog, E. et al. Transcriptomic and morphophysiological evidence for a specialized human cortical GABAergic cell type. Nat. Neurosci. 21, 1185–1195 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  52. Halgren, M. et al. The generation and propagation of the human alpha rhythm. Proc. Natl Acad. Sci. USA 116, 23772–23782 (2019).

    CAS  PubMed  PubMed Central  Google Scholar 

  53. Beaulieu-Laroche, L. et al. Enhanced dendritic compartmentalization in human cortical neurons. Cell 175, 643–651 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  54. Shapson-Coe, A. et al. A connectomic study of a petascale fragment of human cerebral cortex. Preprint at https://www.biorxiv.org/content/10.1101/2021.05.29.446289v1 (2021).

  55. Halgren, E. et al. Processing stages underlying word recognition in the anteroventral temporal lobe. Neuroimage 30, 1401–1413 (2006).

    PubMed  Google Scholar 

  56. Yang, J. C. et al. Microscale dynamics of electrophysiological markers of epilepsy. Clin. Neurophysiol. 132, 2916–2931 (2021).

    PubMed  Google Scholar 

  57. Mosher, C. P. et al. Cellular classes in the human brain revealed in vivo by heartbeat-related modulation of the extracellular action potential waveform. Cell Rep. 30, 3536–3551 (2020).

    CAS  PubMed  PubMed Central  Google Scholar 

  58. Musk, E. & Neuralink. An integrated brain–machine interface platform with thousands of channels. Preprint at https://www.biorxiv.org/content/10.1101/703801v4 (2019).

  59. Dutta, B. et al. The Neuropixels probe: a CMOS based integrated microsystems platform for neuroscience and brain–computer interfaces. 2019 IEEE International Electron Devices Meeting. 10.1.1–10.1.4 https://doi.org/10.1109/IEDM19573.2019.8993611 (2019).

  60. Koch, C. & Jones, A. Big science, team science, and open science for neuroscience. Neuron 92, 612–616 (2016).

    CAS  PubMed  Google Scholar 

  61. Ascoli, G. A. Mobilizing the base of neuroscience data: the case of neuronal morphologies. Nat. Rev. Neurosci. 7, 318–324 (2006).

    CAS  PubMed  Google Scholar 

  62. Ascoli, G. A., Donohue, D. E. & Halavi, M. NeuroMorpho.Org: a central resource for neuronal morphologies. J. Neurosci. 27, 9247–9251 (2007).

    CAS  PubMed  PubMed Central  Google Scholar 

  63. Tóth, K. et al. Hyperexcitability of the network contributes to synchronization processes in the human epileptic neocortex. J. Physiol. 596, 317–342 (2018).

    PubMed  Google Scholar 

  64. Varga, C., Tamas, G., Barzo, P., Olah, S. & Somogyi, P. Molecular and electrophysiological characterization of GABAergic interneurons expressing the transcription factor COUP-TFII in the adult human temporal cortex. Cereb. Cortex 25, 4430–4449 (2015).

    PubMed  PubMed Central  Google Scholar 

  65. Westover, M. B. et al. Real-time segmentation of burst suppression patterns in critical care EEG monitoring. J. Neurosci. Methods 219, 131–141 (2013).

    PubMed Central  Google Scholar 

  66. Patel, S. R. et al. Studying task-related activity of individual neurons in the human brain. Nat. Protoc. 8, 949–957 (2013).

    CAS  PubMed  Google Scholar 

  67. Reuter, M., Rosas, H. D. & Fischl, B. Highly accurate inverse consistent registration: a robust approach. Neuroimage 53, 1181–1196 (2010).

    PubMed  Google Scholar 

  68. Reuter, M., Schmansky, N. J., Rosas, H. D. & Fischl, B. Within-subject template estimation for unbiased longitudinal image analysis. Neuroimage 61, 1402–1418 (2012).

    PubMed  Google Scholar 

  69. Dykstra, A. R. et al. Individualized localization and cortical surface-based registration of intracranial electrodes. Neuroimage 59, 3563–3570 (2012).

    PubMed  Google Scholar 

  70. Holmes, C. J. et al. Enhancement of MR images using registration for signal averaging. J. Comput. Assist. Tomogr. 22, 324–333 (1998).

    CAS  PubMed  Google Scholar 

  71. Felsenstein, O. et al. Multi-Modal Neuroimaging Analysis and Visualization Tool (MMVT). Preprint at https://arxiv.org/abs/1912.10079 (2019).

  72. Lancaster, J. L. et al. Automated regional behavioral analysis for human brain images. Front. Neuroinform. 6, 23 (2012).

  73. Lancaster, J. L. et al. Automated analysis of fundamental features of brain structures. Neuroinformatics 9, 371–380 (2011).

    PubMed  PubMed Central  Google Scholar 

  74. Lancaster, J. L. et al. Anatomical global spatial normalization. Neuroinformatics 8, 171–182 (2010).

    PubMed  PubMed Central  Google Scholar 

  75. Felsenstein, O. & Peled, N. Multi-Modality Visualization Tool (MMVT). https://doi.org/10.5281/zenodo.438343 (2017).

  76. Postelnicu, G., Zöllei, L. & Fischl, B. Combined volumetric and surface registration. IEEE Trans. Med. Imaging 28, 508–522 (2009).

    PubMed  Google Scholar 

  77. Mora Lopez, C. et al. A neural probe with up to 966 electrodes and up to 384 configurable channels in 0.13 μm SOI CMOS. IEEE Trans. Biomed. Circuits Syst. 11, 510–522 (2017).

    PubMed  Google Scholar 

  78. Mian, M. K. et al. Encoding of rules by neurons in the human dorsolateral prefrontal cortex. Cereb. Cortex 24, 807–816 (2014).

    PubMed  Google Scholar 

  79. Dickey, C. W. et al. Travelling spindles create necessary conditions for spike-timing-dependent plasticity in humans. Nat. Commun. 12, 1027 (2021).

    CAS  PubMed  PubMed Central  Google Scholar 

  80. Snyder, A. C., Morais, M. J. & Smith, M. A. Dynamics of excitatory and inhibitory networks are differentially altered by selective attention. J. Neurophysiol. 116, 1807–1820 (2016).

    PubMed  PubMed Central  Google Scholar 

  81. Salami, P., Borzello, M., Kramer, M. A., Westover, M. B. & Cash, S. S. Quantification of seizure termination patterns reveals limited pathways to seizure end. Preprint at https://www.medrxiv.org/content/10.1101/2021.03.03.21252789v1 (2021).

  82. Ganji, M. et al. Development and translation of PEDOT:PSS microelectrodes for intraoperative monitoring. Adv. Funct. Mater. 28, 1700232 (2017).

    Google Scholar 

Download references

Acknowledgements

We would like to thank Y. Chou, A. Tripp, F. Minidio, A. Zhang and A. O’Donnell for help in data collection. We would like to especially thank the patients for their willingness to participate in this research. This research was supported by the ECOR and K24-NS088568 (to S.S.C.) and the Tiny Blue Dot Foundation (to S.S.C. and A.C.P.) and NIH grant U01NS121616 (to Z.M.W.). This research was also supported by NIH NINDS BRAIN R01NS11662301 (to K.V.S.), NIH NIDCD R01DC01403406 (to K.V.S.), the Simons Foundation (543045, to K.V.S.) and the Howard Hughes Medical Institute at Stanford University (to K.V.S., S.D.S. and E.M.T.). S.D.S. was supported by the A. P. Giannini Foundation and the Wu Tsai Neurosciences Institute Interdisciplinary Scholars Fellowship and holds a Career Award at the Scientific Interface from the Burroughs Wellcome Fund. E.M.T. is additionally funded by the Brain and Behavior Research Foundation and the Grossman Institute. The views and conclusions contained in this document are those of the authors and do not represent the official policies, either expressed or implied, of the funding sources. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Author information

Authors and Affiliations

Authors

Contributions

The experiment was conceived by S.S.C., Y.K., A.C.P., Z.W., K.V.S., L.R.H., E.M.T. and S.D.S. Z.W. and R.M.R. performed the surgeries and placed the arrays, and A.C.P., D.J.S., M.M., A.K. and Y.K. collected the data and did first-pass analysis. A.C.P., Y.K. and S.S.C. prepared the first manuscript draft. A.C.P. and Y.K. analyzed all the data for the final results, and A.C.P. prepared all the figures. All authors revised the manuscript. B.D., M.W. and E.M.T. conceived of and advanced the production of the thicker custom Neuropixels probes used in the study. All the authors edited the manuscript.

Corresponding authors

Correspondence to Angelique C. Paulk, Ziv M. Williams or Sydney S. Cash.

Ethics declarations

Competing interests

K.V.S. is a consultant to Neuralink Corporation and CTRL-Labs, Inc. (now part of the Facebook Reality Labs division of Facebook) and is on the scientific advisory boards of Mind-X, Inc., Inscopix, Inc. and Heal, Inc. S.D.S. is a scientific advisor to Nēsos Corporation. The MGH Translational Research Center has clinical research support agreements with Neuralink, Paradromics and Synchron, for which S.S.C. and L.R.H. provide consultative input. None of these entities listed is involved with this research or the Neuropixels device. B.D. and M.W. are employees of IMEC, a non-profit semiconductor research and development organization that manufactures, sells and distributes the Neuropixels probes, at cost, to the research community. The remaining authors declare no competing interests.

Peer review

Peer review information

Nature Neuroscience thanks Ueli Rutishauser and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Recording challenges and lessons learned.

Recording challenges and lessons learned. a. Diagram of the different combinations of ground and referencing and connections that improved the electrophysiological signal or degraded the signal with the indicated ground and reference contacts (including the external and internal reference). Each circled number indicates a combination tested. A major step in reducing noise levels was to separate the ground and reference, with a single separate wire going to the grounding pads on one side and another wire going to a grounding pad on the other side of the Neuropixels probe, pictured here. We used the external reference tied to a sterile MedTronic grounding wire with a needle which, when placed in the scalp or CSF, improved the signal and reduced 60 Hz noise. Common 60 Hz noise and other noise decreased considerably if the anesthesia IV pump was unplugged from the wall and was run on battery during the recordings. Otherwise, turning off lights or other sources of noise (BOVIE cautery machine, AlphaOmega recording system, etc.) had no noticeable effect on noise. b. Diagram of the connections relative to the participant including the BOVIE pad placed on the leg connected to the IMEC ground via a BNC cable to the NIDAQ PCI chassis BNC. c. Securing the wires and the probe to the cannulae attached to the AlphaOmega manipulator with Tegaderm and Steristrips improved the stability. d. Mechanical stabilization of the probe involved two options, one using the ROSA robot combined with an AlphaOmega manipulator with the Neuropixels probe secured to cannulae. e. The second option involved securing the Neuropixels probe to a sterile syringe using Steristrips. f. Then, the syringe is attached to a 3-axis manipulator mounted on a Greenberg retractor over the craniotomy. g. Every Neuropixels probe is documented and checked several times during and after the procedure both via software (in the OpenEphys or SpikeGLX software) and a quick saline test is done in the side sterile table. h. Each probe is photographed before, during, and afterward in the OR to determine if they are intact. Regular checks of the connections are done as the electrode is being moved into place in the sterile field. The electrode is also checked and photographed afterward for whether it is intact and connected. i. In two cases, 60 Hz noise was reduced by tying a ground to a spare BOVIE pad placed on the thigh of the patient under the sterile drape (as shown in b), with the traces shown with the lower left LFP trace showing the signal grounded to the BOVIE versus grounded to a scalp needle electrode. j. Examples of considerable noise in the LFP and high frequency range (AP range) when crossing the ground and reference together. k. We did test using the internal reference on the Neuropixels probe and found the noise increased considerably in the two cases we attempted the switch. l. The placement of the sterile ground and reference leads made a difference. Ground and reference in the scalp had an improved signal. Placing the ground (but not the reference) in the saline in the craniotomy caused the LFP signal to saturate and degraded the signal.

Extended Data Fig. 2 Putative Neuropixels probe location based on photographic evidence and identification of intracranial access to the cortical surface.

Putative Neuropixels probe location based on photographic evidence and identification of intracranial access to the cortical surface. a. preoperative MRI (Pt. 01) (displayed with Mango software10–12) with an overlay of the postoperative CT including the DBS leads and burr hole used to implant the DBS leads. As the Neuropixels probe was inserted through the burr hole, there was only a limited number of areas and approaches using angles in three dimensions to insert the Neuropixels probe. b. Taking measures of the burr hole (upper right in a), we reconstructed the burr hole relative to the participant’s brain (produced using FreeSurfer and MMVT5–9, to identify the possible locations for the Neuropixels probe insertion taking into account the size of the 3D structure of the headstage and insertion into a cortical gyrus. Grey surface is the pial surface and white reconstructed volume is the white matter for the participant. c-d. The resultant putative location of the Neuropixels probe based on these measures in 3D (c) and overlaid on the MRI (d). e-g. Same process but for Pt. 02. h. Photographs of the open craniotomy (Pt. 03) and process of tracing the sulci to identify location of the craniotomy. i. Reconstructed brain with the overlaid trace of the craniotomy photograph of the lateral temporal lobe, with the putative location of the inserted Neuropixels probe shown on the brain. j-k. The resultant putative location of the Neuropixels probe based on the projected photograph to the 3D location on the brain.

Extended Data Fig. 3 Realigning the data relative to heartbeat-induced movement artifact.

Realigning the data relative to heartbeat-induced movement artifact. a. Illustration of evidence of tissue movement relative to the electrode recordings in the LFP (shown in red-blue color scale with the range in µvolts shown in b). This is quantified by manually tracing these ‘band shifts’ using the Blender program, followed by detection of these movements in the LFP and tracking of these movements across channels (white line, second to rightmost plot). b. LFP before (left) and after (right) adjusting for movement effects. c. High frequency (action potential) frequency signal before (left) and after (right) adjusting for movement effects. d. Top row: Kilosort 3.0 registration and alignment alone could not compensate for the drift evident in the detected spike waveforms (left) and the estimated drift spanned hundreds of microns (right). Bottom row: manual alignment (a-b) followed by Kilosort 3.0 sorting resulted in improved spike alignment through time (left) and reduced drift (right).

Extended Data Fig. 4 Cross correlation between the interpolated sorted units and the raw data.

Cross correlation between the interpolated sorted units and the raw data. a. tracked movement from the LFP for the three participants, including zoomed in views in the grey boxes. b. Log power of the tracked movement from LFP for one participant for 0-50 Hz (power curves, left, and spectrogram, right). c. Log power of the tracked movement from LFP for one participant for 0–5 Hz (power curves, left, and spectrogram, right). d. Power of two different frequency bands (1.5 and 0.02 Hz) of the tracked movement from LFP through time for one participant (power curves). e. Moving average calculated velocity (µm/sec) of the movement as tracked in the LFP per participant. f. Left: We could correlate the 12 channels with the waveforms on a per-unit and per channel level and overlay the raw data (black lines) and the sorted movement tracked data (red lines). Right: Approach for cross correlating the movement-corrected neural data and the raw signals by cross correlating the 12 channels of the sorted spikes with the raw data with tracking of individual units along the raw channel data (top, middle, purple dots) based on the cross-correlation values (shown in color map) which was paired with the LFP tracked movement (white lines). g. Mean and median velocity (µm/sec) of the movement as tracked in the LFP per participant. h. Correlation values between sorted units and raw data per unit across 12 channels per unit (Kruskal-Wallis multiple comparisons test; Chi-sq.=12.29; p = 0.0019). Scatter plots are individual unit correlation values per participant. i. Correlation values on the per-channel waveform basis between the interpolated and raw data sets versus the log spike rates per unit. Each dot indicates a unit. j. Mean variance in tracked cross-correlated single units. Asterisk indicates significant differences between each participant in the variance (Kruskal-Wallis multiple comparisons test; Chi-sq.=399.39; p < 0.00001).

Extended Data Fig. 5 Example complex waveforms for six different units (each color-coded set of waveforms) across the data set.

Example complex waveforms for six different units (each color-coded set of waveforms) across the data set. Original waveforms are overlaid relative to the recorded channels, with the grey bars to the right indicating the location of the units along the Neuropixels probe.

Extended Data Fig. 6 PCA with k-means clustering.

PCA with k-means clustering. a. Percent variance explained by each principal component calculated across the first 6 channels per unit. b. Silhouette values for each k-means cluster number for the first 40 principal components after re-running the k-means clustering analyses 20 times (including 1000 iterations). Error bars indicate SEM over the repeated clustering analyses. c. Scatter plot of seven clusters (using k-means clustering) in the first three principal components. Color coding reflects different clusters.

Extended Data Fig. 7 Waveform Features of Units Clustered with WaveMAP per participant.

Waveform Features of Units Clustered with WaveMAP per participant. a. Waveforms per cluster per patient (indicated as Pt. xx). The sorted clusters (with WaveMAP) was performed on the per-patient level which is in contrast with the WaveMAP clustering performed across all three participants (Fig. 3). b. The remaining measures are per patient, showing mean firing rate, peak-trough ratio, spatial spread, and depth violin plots for the different clusters. * indicate significant differences between putative cell or waveform types, Kruskal-Wallis multiple comparison test, p < 0.001. c. Electrode locations relative to the cortical surface and cortical regions.

Extended Data Fig. 8 Waveform measures.

Waveform measures and PCA with k-means clustering. a. Waveform measures on an example, spike waveform including the spike duration, halfwidth, peak-trough ratio, repolarization slope, recovery slope, and amplitude measures. b. Largest waveforms per unit showing positive polarity (grey) and negative polarity (black) waveforms across the data sets. c. Distribution of positive polarity (grey) and negative polarity (black) half-peak width and spike duration (peak-trough duration) of the largest waveform per unit (Pt. 01-03). d. The peak-trough ratio, repolarization slope, recovery slope of the channel with the largest waveform violin plots for the different clustered waveforms using the single channel and multichannel classifications. * indicate significant differences between all putative cell or waveform types, Kruskal-Wallis multiple comparison test, p < 0.001 with post hoc Tukey-Kramer test. Grey letters a-c indicate statistically separable groups. Left: single channel classifications. Middle: multichannel classification based on the principal components and k-means clustering across patients. Right: multichannel classification based on the WaveMAP clustering across patients.

Extended Data Fig. 9 Spatial waveform measures.

a. Spatial spread examples of individual units (averaged across all waveforms attributed to a given unit) across the left and right columns of the Neuropixels probe. Voltage indicated by the greyscale color scheme. Bottom plot: diagram (arrows) demonstrating the spatial mapping calculations, including spatial spread, velocity above (yellow arrow), and velocity below (cyan arrow) the center point (channel with the largest waveform). (Measures shown in Extended Data Fig. 4). b. The spatial spread, depth, velocity above, and velocity below the center point (channel with the largest waveform) violin plots are shown for the different waveform types. * indicate significant differences between all putative cell or waveform types, Kruskal-Wallis multiple comparison test, p < 0.001 with post hoc Tukey-Kramer test. Grey letters a-c indicate statistically separable groups. Left: single channel classifications. Middle: multichannel classification based on the principal components and k-means clustering across patients. Right: multichannel classification based on the WaveMAP clustering across patients.

Extended Data Fig. 10 Single units through time and correlation relationships between units.

Single units through time and correlation relationships between units. a. Left: the raster plots of spike times throughout the recording for the different classes of single units as clustered by WaveMAP (color coded relative to the waveform from the channel with the largest amplitude per unit) as well as the MUA activity (grey). Right: Spike counts binned in 5 second bins throughout the recordings for the three participants. b. Example cross-covariance of the spike times between individual units of different putative classes as clustered by WaveMAP (as titled). The vertical blue line shows the mean per plot.

Supplementary information

Supplementary Information

Supplementary Table 1

Reporting Summary

Supplementary Video 1

Movie of raw and interpolated Neuropixels data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Paulk, A.C., Kfir, Y., Khanna, A.R. et al. Large-scale neural recordings with single neuron resolution using Neuropixels probes in human cortex. Nat Neurosci 25, 252–263 (2022). https://doi.org/10.1038/s41593-021-00997-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41593-021-00997-0

Further reading

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing